summaryrefslogtreecommitdiffstats
path: root/drivers/net
AgeCommit message (Collapse)Author
2010-05-12netxen: remove unnecessary size checksSucheta Chakraborty
NX3031 have 64bit on card memory. Fix the limit check to 64MB and remove unnecessary 128bit read/write check. Signed-off-by: Sucheta Chakraborty <sucheta.chakraborty@qlogic.com> Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-05-12netxen: fix register usageAmit Kumar Salecha
o For NX3031, MSI_MODE, CAPABILITIES_FW and SCRATCHPAD registers are obsolete. These register addresses can be used for different purpose. Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-05-12Merge branch 'master' of ↵David S. Miller
master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6 Conflicts: Documentation/feature-removal-schedule.txt drivers/net/wireless/ath/ar9170/usb.c drivers/scsi/iscsi_tcp.c net/ipv4/ipmr.c
2010-05-11ar9170: fix for driver-core ABI changeStephen Rothwell
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-11Merge branch 'master' of ↵John W. Linville
git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next-2.6 into for-davem Conflicts: drivers/net/wireless/ath/ar9170/main.c
2010-05-11Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6Linus Torvalds
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: net: Fix FDDI and TR config checks in ipv4 arp and LLC. IPv4: unresolved multicast route cleanup mac80211: remove association work when processing deauth request ar9170: wait for asynchronous firmware loading ipv4: udp: fix short packet and bad checksum logging phy: Fix initialization in micrel driver. sctp: Fix a race between ICMP protocol unreachable and connect() veth: Dont kfree_skb() after dev_forward_skb() IPv6: fix IPV6_RECVERR handling of locally-generated errors net/gianfar: drop recycled skbs on MTU change iwlwifi: work around passive scan issue
2010-05-10Merge branch 'master' of ↵David S. Miller
git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-2.6
2010-05-10wireless: depends on NETRandy Dunlap
When CONFIG_NET is disabled, the attempt to build wext-priv.c fails with: net/wireless/wext-priv.c: In function 'ioctl_private_call': net/wireless/wext-priv.c:207: error: implicit declaration of function 'call_commit_handler' Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-10wireless: rt2x00: rt2800usb: replace X by xXose Vazquez Perez
s/X/x Signed-off-by: Xose Vazquez Perez <xose.vazquez@gmail.com> Acked-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-10rt2x00: Clean up generic procedures on descriptor writing.Gertjan van Wingerde
With a little bit of restructuring it isn't necessary to have special cases in rt2x00queue_write_tx_descriptor for writing the descriptor for beacons. Simply split off the kicking of the TX queue to a separate function with is only called for non-beacons. Signed-off-by: Gertjan van Wingerde <gwingerde@gmail.com> Acked-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-10rt2x00: Fix beaconing on rt2800.Gertjan van Wingerde
According to the Ralink vendor driver for rt2800 we don't need a full TXD for a beacon but just a TXWI in front of the actual beacon. Fix the rt2800pci and rt2800usb beaconing code accordingly. Signed-off-by: Gertjan van Wingerde <gwingerde@gmail.com> Acked-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-10rt2x00: provide beacon's txdesc to write_beacon callback function.Gertjan van Wingerde
Preparation to fix rt2800 beaconing. Signed-off-by: Gertjan van Wingerde <gwingerde@gmail.com> Acked-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-10rt2x00: Clean up all driver's kick_tx_queue callback functions.Gertjan van Wingerde
All of the driver's kick_tx_queue callback functions treat the TX queue for beacons in a special manner. Clean this up by integrating the kicking of the beacon queue into the write_beacon callback function, and let the generic code no longer call the kick_tx_queue callback function when updating the beacon. Signed-off-by: Gertjan van Wingerde <gwingerde@gmail.com> Acked-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-10rt2x00: Factor out RXWI processing to common rt2800 code.Gertjan van Wingerde
RXWI processing is exactly the same for rt2800pci and rt2800usb, so make it common code. Signed-off-by: Gertjan van Wingerde <gwingerde@gmail.com> Acked-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-10rt2x00: Factor out TXWI writing to common rt2800 code.Gertjan van Wingerde
TXWI writing is exactly the same for rt2800pci and rt2800usb, so make it common code. Signed-off-by: Gertjan van Wingerde <gwingerde@gmail.com> Acked-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-10rt2x00: Don't check whether hardware crypto is enabled when reading RXD.Gertjan van Wingerde
We should simply follow what the hardware told us it has done. Signed-off-by: Gertjan van Wingerde <gwingerde@gmail.com> Acked-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-10rt2x00: Clean up rt2800usb.h.Gertjan van Wingerde
Remove unused RXD_DESC_SIZE define and remove duplicated RXWI definitions from rt2800.h. Signed-off-by: Gertjan van Wingerde <gwingerde@gmail.com> Acked-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-10rt2x00: Fix setting of txdesc->length field.Gertjan van Wingerde
We should take the stripping of the IV into account for the txdesc->length field. Signed-off-by: Gertjan van Wingerde <gwingerde@gmail.com> Acked-by: Ivo van Doorn <IvDoorn@gmail.com> Tested-by: Pavel Roskin <proski@gnu.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-10ath5k: several off by one range checksDan Carpenter
There are several places that use > ARRAY_SIZE() instead of >= ARRAY_SIZE(). Signed-off-by: Dan Carpenter <error27@gmail.com> Acked-by: Bob Copeland <me@bobcopeland.com> Acked-by: Bruno Randolf <br1@einfach.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-10ath9k/htc_drv_main: off by one errorDan Carpenter
I changed "> ATH9K_HTC_MAX_TID" to ">= ATH9K_HTC_MAX_TID" to avoid a potential overflow. Signed-off-by: Dan Carpenter <error27@gmail.com> Acked-by: Sujith <Sujith.Manoharan@atheros.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-10ath9k/htc_drv_main: null dereference typoDan Carpenter
This is a stray null dereference. We initialize "ista" properly later on. Signed-off-by: Dan Carpenter <error27@gmail.com> Acked-by: Sujith <Sujith.Manoharan@atheros.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-10iwlwifi: remove stray mutex_unlock()Dan Carpenter
This mutex_unlock() has been here from the initial commit, but as nearly as I can tell, there isn't a reason for it. Signed-off-by: Dan Carpenter <error27@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-10ath9k_hw: enable PCIe low power mode for AR9003Luis R. Rodriguez
Cc: Paul Shaw <paul.shaw@atheros.com> Cc: Don Breslin <don.breslin@atheros.com> Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-10rtl8180: change PCI DMA mask to DMA_BIT_MASK(32)John W. Linville
From the original report: "I had problems to get my rtl8185 PCI card running on Sparc64: I always got an error about "No suitable DMA available" followed by an error that no device could be detected. When comparing the rtl8180 driver to others I noticed that others are mostly using DMA_BIT_MASK so I changed the custom mask to DMA_BIT_MASK(32) which fixed my issue." Reported-by: Tiziano Müller <tm@dev-zero.ch> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-10net: trans_start cleanupsEric Dumazet
Now that core network takes care of trans_start updates, dont do it in drivers themselves, if possible. Drivers can avoid one cache miss (on dev->trans_start) in their start_xmit() handler. Exceptions are NETIF_F_LLTX drivers Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-05-07rt2x00: rt2800: use correct txop value in tx descriptorHelmut Schaa
rt2800 devices use a different enumeration to specify what IFS values should be used on frame transmission compared to the other rt2x00 devices. Hence, create a new enum called txop that contains the valid values. Furthermore use the appropriate txop values as found in the ralink drivers: - TXOP_BACKOFF for management frames - TXOP_SIFS for subsequent fragments in a burst - TXOP_HTTXOP for all data frames Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com> Acked-by: Ivo van Doorn <IvDoorn@gmail.com> Acked-by: Gertjan van Wingerde <gwingerde@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07wl1271: Reduce PSM entry hang over period from 128 => 1 msJuuso Oikarinen
Currently, we configure a 128ms hang over period for the PSM entry (the firmware will remain active for 128ms after sending the null func for PSM and getting an ack for it.) This is a huge power consumption issue, and appears unnecessary. So, configure the value to 1 ms. Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com> Reviewed-by: Janne Ylalehto <janne.ylalehto@nokia.com> Signed-off-by: Luciano Coelho <luciano.coelho@nokia.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07wl1271: Increase timeout for command event waitingTeemu Paasikivi
Incresed the timeout value for command complete event waiting from 100 ms to 750 ms. In some rare cases it can take about 600 ms before complete event for join command is received. This is most propably caused by the firmware being busy with scanning related activities. Signed-off-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com> Reviewed-by: Juuso Oikarinen <juuso.oikarinen@nokia.com> Signed-off-by: Luciano Coelho <luciano.coelho@nokia.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07wl1271: Fix to join and channel number handlingJuuso Oikarinen
This patch changes the way JOIN's are performed, and channel numbers updated. The reason for this is that the firmware JOIN command clears WPA(2) key material, and if done while associated to a WPA(2) secured AP, will render the data-path unusable. While the channel is not usually changed while associated (and currently we could not even support something like that), after performing a scan operation while associated, mac80211 will re-set the current channel to the driver. This caused our problem. Also, the mac80211 is assuming that the driver channel configuration remains persistent over periods of IDLE. Therefore remove channel resetting to zero from the unjoin function. Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com> Reviewed-by: Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com> Signed-off-by: Luciano Coelho <luciano.coelho@nokia.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07wl1271: Fix 32 bit register read related endiannes bugJuuso Oikarinen
Reading single registers did not pay attention to data endianness. This patch fix that. Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com> Reviewed-by: Luciano Coelho <luciano.coelho@nokia.com> Signed-off-by: Luciano Coelho <luciano.coelho@nokia.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07wl1271: Add sysfs file to retrieve HW PG-version and ROM-versionJuuso Oikarinen
This patch reads the HW PG version (along with a ROM-version, embedded in the same value) from the wl1271 hardware and publishes the value in a sysfs -file. Signed-off-by: Juuso Oikarinen <juuso.oikarinen@nokia.com> Reviewed-by: Luciano Coelho <luciano.coelho@nokia.com> Signed-off-by: Luciano Coelho <luciano.coelho@nokia.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07ath9k_htc: Handle IDLE LED properlySujith
Switch LED off/on when handling CONF_CHANGE_IDLE. Not doing this would leave the radio LED on even though the chip would be in full sleep mode. Signed-off-by: Sujith <Sujith.Manoharan@atheros.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07ath9k_hw: Update initvals for AR9003 for xb113Luis R. Rodriguez
Generated using the new shiny intivals-tool [1]: initvals -w -f ar9003 > ar9003_initvals.h The respective checksums are: 0x000000005a76829d ar9300_2p0_radio_postamble 0x000000009d90cb74 ar9300Modes_lowest_ob_db_tx_gain_table_2p0 0x00000000e0bc2c84 ar9300Modes_fast_clock_2p0 0x00000000852fca34 ar9300_2p0_radio_core 0x0000000000000000 ar9300Common_rx_gain_table_merlin_2p0 0x0000000078658fb5 ar9300_2p0_mac_postamble 0x0000000023235333 ar9300_2p0_soc_postamble 0x0000000054d41904 ar9200_merlin_2p0_radio_core 0x00000000618455d4 ar9300_2p0_baseband_postamble 0x000000009aa590a4 ar9300_2p0_baseband_core 0x000000004783d946 ar9300Modes_high_power_tx_gain_table_2p0 0x000000006681db44 ar9300Modes_high_ob_db_tx_gain_table_2p0 0x000000001f318700 ar9300Common_rx_gain_table_2p0 0x000000009990cb74 ar9300Modes_low_ob_db_tx_gain_table_2p0 0x00000000c9d66d40 ar9300_2p0_mac_core 0x0000000039139500 ar9300Common_wo_xlna_rx_gain_table_2p0 0x00000000a0c54980 ar9300_2p0_soc_preamble 0x00000000292e2544 ar9300PciePhy_pll_on_clkreq_disable_L1_2p0 0x000000002d3e2544 ar9300PciePhy_clkreq_enable_L1_2p0 0x00000000293e2544 ar9300PciePhy_clkreq_disable_L1_2p0 [1] http://wireless.kernel.org/en/users/Drivers/ath9k_hw/initvals-tool Cc: Tom Hammel <thammel@atheros.com> Cc: Enis Akay <Enis.Akay@Atheros.com> Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07ath9k_common: drop incomming frames with an invalid hardware rateLuis R. Rodriguez
ath9k_common (used by ath9k and ath9k_htc) trusts the frames blessed by hardware as OK are infact correct even if the rate seen by the driver is unrecognized. ath9k_common just treats these frames in mac80211 as frames as frames under 1 mbps rate. It seems this might not be the best thing to do as other parts of the frame might not be valid so just drop these frames for now. Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07ath9k_common: move the rate status setting into ath9k_process_rate()Luis R. Rodriguez
This has no real functional change, this just moves the setting the the mac80211 rate index into ath9k_process_rate(). This allows us to eventually make ath9k_process_rate() return a negative value in case we have detected a specific case rate situation which should have been ignored. Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07rtl8180: add software-based support for IBSS modeJohn W. Linville
Device documentation suggests that hardware support for beaconing is available. But I implemented software-based beacon generation as an experiment and it seems better to have that working now rather than waiting for something better to materialize. Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07rtl8180: assign sequence numbers in the driverJohn W. Linville
This is a step towards support for beaconing modes of operation. Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07rt2x00: Fix RF3052 channel initializationIvo van Doorn
Update channel initialization for the RF3052 chipset. According to the Ralink drivers, the rt3x array must be used for this chipset, rather then the rt2x array. Furthermore RF3052 supports the 5GHz band, extend the rt3x array with the 5GHz channels, and use them for the RF3052 chip. Signed-off-by: Ivo van Doorn <IvDoorn@gmail.com> Acked-by: Gertjan van Wingerde <gwingerde@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07rt2x00: rt2800: don't overwrite SIFS values on erp changesHelmut Schaa
The SIFS value is a constant and doesn't need to be updated on erp changes. Furthermore the code used 10us for both, the OFDM SIFS and CCK SIFS time which broke CTS protected 11g connections (see patch "rt2x00: rt2800: update initial SIFS values" for details). Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com> Acked-by: Ivo van Doorn <IvDoorn@gmail.com> Acked-by: Gertjan van Wingerde <gwingerde@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07rt2x00: rt2800: update initial SIFS valuesHelmut Schaa
Currently the CCK and OFDM SIFS value is set to 32us. This value is neither used by the Ralink driver nor specified in 802.11. Instead of using 10us for CCK SIFS (as defined in 802.11) use 16us like in the Ralink drivers. And indeed using a SIFS value of 10us breaks connectivity with 11g + CTS protected connections. Add a comment to the code why we don't use 10us for CCK SIFS value. The OFDM SIFS value is set to 16us (as defined in 802.11 and also used by the Ralink drivers). Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com> Acked-by: Ivo van Doorn <IvDoorn@gmail.com> Acked-by: Gertjan van Wingerde <gwingerde@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07ath9k_htc: Fix beaconing in IBSS modeSujith
The current way of managing beaconing in ad-hoc mode has a subtle race - the beacon obtained from mac80211 is freed in the SWBA handler rather than the TX completion routine. But transmission of beacons goes through the normal SKB queue maintained in hif_usb, leading to a situation where __skb_dequeue() in the TX completion handler goes kaput. Fix this by simply getting a beacon from mac80211 for every SWBA and free it in its completion routine. Signed-off-by: Sujith <Sujith.Manoharan@atheros.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07mac80211: improve HT channel handlingJohannes Berg
Currently, when one interface switches HT mode, all others will follow along. This is clearly undesirable, since the new one might switch to no-HT while another one is operating in HT. Address this issue by keeping track of the HT mode per interface, and allowing only changes that are compatible, i.e. switching into HT40+ is not possible when another interface is in HT40-, in that case the second one needs to fall back to HT20. Also, to allow drivers to know what's going on, store the per-interface HT mode (channel type) in the virtual interface's bss_conf. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07cfg80211/mac80211: better channel handlingJohannes Berg
Currently (all tested with hwsim) you can do stupid things like setting up an AP on a certain channel, then adding another virtual interface and making that associate on another channel -- this will make the beaconing to move channel but obviously without the necessary IEs data update. In order to improve this situation, first make the configuration APIs (cfg80211 and nl80211) aware of multi-channel operation -- we'll eventually need that in the future anyway. There's one userland API change and one API addition. The API change is that now SET_WIPHY must be called with virtual interface index rather than only wiphy index in order to take effect for that interface -- luckily all current users (hostapd) do that. For monitor interfaces, the old setting is preserved, but monitors are always slaved to other devices anyway so no guarantees. The second userland API change is the introduction of a per virtual interface SET_CHANNEL command, that hostapd should use going forward to make it easier to understand what's going on (it can automatically detect a kernel with this command). Other than mac80211, no existing cfg80211 drivers are affected by this change because they only allow a single virtual interface. mac80211, however, now needs to be aware that the channel settings are per interface now, and needs to disallow (for now) real multi-channel operation, which is another important part of this patch. One of the immediate benefits is that you can now start hostapd to operate on a hardware that already has a connection on another virtual interface, as long as you specify the same channel. Note that two things are left unhandled (this is an improvement -- not a complete fix): * different HT/no-HT modes currently you could start an HT AP and then connect to a non-HT network on the same channel which would configure the hardware for no HT; that can be fixed fairly easily * CSA An AP we're connected to on a virtual interface might indicate switching channels, and in that case we would follow it, regardless of how many other interfaces are operating; this requires more effort to fix but is pretty rare after all Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07orinoco: refactor xmit pathDavid Kilroy
... so orinoco_usb can share some common functionality. Handle 802.2 encapsulation and MIC calculation in that function. The 802.3 header is prepended to the SKB. The calculated MIC is written to a specified buffer. Also modify the transmit control word that will be passed onto the hardware to specify whether the MIC is present, and the key used. Signed-off-by: David Kilroy <kilroyd@googlemail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07ath9k: fix another source of corrupt framesFelix Fietkau
Atheros hardware supports receiving frames that span multiple descriptors and buffers. In this case, the rx status of every descriptor except for the last one is invalid and may contain random data. Because the driver does not support this, it needs to drop such frames. Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07ar9170usb: remove deprecated aggregation codeChristian Lamparter
This patch removes the incomplete AMPDU implementation in ar9170usb. The code in question is: * too big and complex (more than 550 SLOC.) This is enough to qualify for a new separate code file! * unbalanced quantity & quality over-engineered areas like: * xmit scheduling and queuing frames for multiple HT peers * redundant frame sorting are confronted by gaping holes: * accurate transmission feedback * firmware error-handling and device reset * HT rate control algorithm * error-prone Since its inclusion, hardly anything was done to fix any of the outlined flaws from the initial commit message. => This also indicates poor maintainability. * relies heavily on several spinlocks. As a result of this shortcomings, the code is slow and does not even support the most basic 11n requirement: HT station mode. Therefore, I request to purge my heap of **** from the kernel: "ar9170: implement transmit aggregation". The next item on the agenda is: (re-)start from scratch with an adequate design to accommodate the special requirements and features of the available frameworks and tools. Signed-off-by: Christian Lamparter <chunkeey@googlemail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-07ar9170: wait for asynchronous firmware loadingChristian Lamparter
This patch fixes a regression introduced by the following patch: "ar9170: load firmware asynchronously" When we kick off a firmware loading request and then unbind, or disconnect the usb device right away, we get into trouble: > ------------[ cut here ]------------ > WARNING: at lib/kref.c:44 kref_get+0x1c/0x20() > Hardware name: 18666GU > Modules linked in: ar9170usb [...] > Pid: 6588, comm: firmware/ar9170 Not tainted 2.6.34-rc5-wl #43 > Call Trace: > [<c102b05e>] ? warn_slowpath_common+0x6e/0xb0 > [<c117c93c>] ? kref_get+0x1c/0x20 > [<c102b0b3>] ? warn_slowpath_null+0x13/0x20 > [<c117c93c>] ? kref_get+0x1c/0x20 > [<c117bb2f>] ? kobject_get+0xf/0x20 > [<c124d630>] ? get_device+0x10/0x20 > [<c124e5a0>] ? device_add+0x60/0x530 > [<c117b8b5>] ? kobject_init+0x25/0xa0 > [<c12569f9>] ? _request_firmware+0x139/0x3e0 > [<c1256cc0>] ? request_firmware_work_func+0x20/0x70 > [<c1256ca0>] ? request_firmware_work_func+0x0/0x70 > [<c103ff24>] ? kthread+0x74/0x80 > [<c103feb0>] ? kthread+0x0/0x80 > [<c1003136>] ? kernel_thread_helper+0x6/0x10 >---[ end trace 2d50bd818f64a1b7 ]--- - followed by a random Oops - Avoid that by waiting for the firmware loading to finish (whether successfully or not) before the unbind in ar9170_usb_disconnect. Reported-by: Johannes Berg <johannes@sipsolutions.net> Bug-fixed-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: Christian Lamparter <chunkeey@googlemail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2010-05-06bnx2: Add prefetches to rx path.Michael Chan
Add prefetches of the skb and the next rx descriptor to speed up rx path. Use prefetchw() for the skb [suggested by Eric Dumazet]. The rx descriptor is in skb->data which is mapped for streaming mode DMA. Eric Dumazet pointed out that we should not prefetch the data before dma_sync. So we prefetch only if dma_sync is no_op on the system. Signed-off-by: Michael Chan <mchan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-05-06bnx2: Add GRO support.Michael Chan
And turn on NETIF_F_GRO by default [requested by DaveM]. Signed-off-by: Michael Chan <mchan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-05-06tehuti: Remove unnecessary memset of netdev private dataTobias Klauser
The memory for the private data is allocated using kzalloc in alloc_etherdev (or alloc_netdev_mq respectively) so there is no need to set it to 0 again. Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Signed-off-by: David S. Miller <davem@davemloft.net>